Modeling and Control of AG Mill Energy Consumption Based on Ore Hardness Distribution in a Large-Scale Iron Ore Plant

Abstract

The performance of autogenous (AG) grinding circuits is highly sensitive to variations in ore hardness, particularly in dry processing operations with limited buffering capacity. This study investigates the influence of ore hardness variability on the operational behavior of the AG mill in Line 3 of the Gole-Gohar Iron Ore Concentration Plant. A total of 82 feed samples were collected and analyzed using the SAG Power Index (SPI) test to quantify ore hardness. The SPI results ranged from 48 to 236 min, revealing substantial heterogeneity in the run-of-mine feed. Correlation analyses demonstrated that increasing ore hardness results in reduced mill feed rate and increased specific energy consumption, primarily due to extended residence time and lower breakage efficiency. A linear regression model was developed to predict the mill’s specific energy demand as a function of SPI, providing a practical tool for energy forecasting and operational optimization. Furthermore, hardness-based ore classification and blending strategies were designed to mitigate the adverse impact of hard feed on mill performance. These homogenization approaches, supported by block model data and SPI testing, offer a cost-effective means to enhance process stability and energy efficiency. Overall, the findings underscore the importance of incorporating quantitative hardness characterization into comminution circuit design, control systems, and mine-to-mill planning frameworks.